Haemoglobin predicts survival in patients with chronic heart failure: a substudy of the ELITE II trial.

AIMS The correction of anaemia in chronic heart failure (CHF) has been suggested to be associated with an improvement in symptoms and cardiac function. We aimed to investigate the relationship between the concentration of haemoglobin and survival in CHF. METHODS AND RESULTS We analysed haemoglobin concentrations in 3044 patients recruited in the Evaluation of Losartan In The Elderly (ELITE II) trial. Patients of mean age 71.5 +/- 6.8 years (+/-SD) and New York Heart Association (NYHA) class 2.5 +/- 0.6 were enrolled from June 1997 to May 1998 and followed-up for survival (range 1-780 days, median 551). In univariate analysis, age, NYHA class, serum creatinine, left ventricular ejection fraction (all P<0.0001) and sex (P=0.046) all predicted survival. Haemoglobin as a continuous variable for all patients was not a significant prognostic marker (P=0.26). However, sub-dividing patients according to 1.0 g/dL increments of haemoglobin revealed that the survival relationship was non-linear. The results from the polynomial regression suggest that the optimal interval is a symmetric one centred around 14.5 g/dL. This was independent of age, sex, NYHA class, left ventricular ejection fraction, creatinine, co-existing chronic obstructive pulmonary disease and treatment allocation (P<0.001). There was a minor fall in plasma haemoglobin at the 12-month follow-up (mean change for all patients 0.3 +/- 2.2 g/dL, P<0.0001), with no difference between captopril and losartan groups (P>0.3). CONCLUSION Haemoglobin is an independent predictor of mortality in CHF patients, with anaemic and polycythaemic patients having the worst survival.